Welcome,

In this course, we aim to specialize in artificial intelligence by working on 7 Machine Learning Projects and 7 Deep Learning Projects (14 AI Projects in total) at various levels (easy - medium - hard). Before starting the course, you should have basic Python knowledge. Our aim in this course is to turn real-life problems into projects and then solve them using latest versions of artificial intelligence algorithms (machine learning algortihms and deep learning algorithms) and Python.

We will carry out some of our projects using machine learning and some using deep learning algorithms. In this way, you will have a general perspective on artificial intelligence. When you complete the projects in our course, you will get a clear understanding of the basic working principles of Machine Learning software and Deep Learning algorithms and the difference between them.

In our course, we will use well known datasets that are widely used by high level education about Machine Learning as well as custom datasets. By doing our projects, you will master artificial intelligence concepts as well as learn these famous datasets. After completing the course, you will be able to easily produce solutions to the problems that you may encounter in real life.

In our Machine Learning Projects we will use Scikit-Learn Python library. In our Deep Learning Projects we will use Tensorflow and Keras libraries.

The course is composed of 14 Artificial Intelligence Projects - Machine Learning Projects and Deep Learning Projects:

– Project #1: House Price Prediction using Machine Learning

In this project we will build a artificial intelligence model that predicts house prices using sklearn multiple linear regression algortihm.

– Project #2: Salary Calculation using Machine Learning

It is a tedious work to calculate each employee’s salary according to employee’s experience level. In this project we are going to build a machine learning model for exact calculation of employee salaries. Since most of salary values are non-linear, a simple linear function can not be used for this calculation process. Generally most of the companies have polynomial salary values for their employees. Therefore we will use polynomial linear regression algorithm for solution here.

– Project #3: Handwritten Digit Recognition using Multiple Machine Learning Models

In this Project, we will implement a software that recognizes and makes sense of the objects in the photograph by using multiple Machine Learning Models together. Thanks to this project, you will see how you can combine machine learning models and combine several models to solve complex problems. You will have solved a problem that can be used in daily life (recognition of a handwritten text by a computer) using Artificial intelligence (AI).

– Project #4: Advanced Customer Segmentation using Machine Learning

In this project, we will use a new and advanced segmentation library developed by the Massachusetts Institute of Technology (MIT). The customer data in our Customer Segmentation project, which is included in the entry and intermediate level projects, was simple and the K-Means clustering algorithm was sufficient for segmentation. But life is not that simple! When you have complex customer data, if you do clustering with K-Means, you may get erroneous results! Since the customer data in this project is complex data (both numeric and categorical) just like in real life, here we will use a special unsupervised learning algorithm instead of a standard model and divide our 2000 customers into groups with the latest artificial intelligence algorithms.

– Project #5: IMDB Sentiment Analysis Using NLP (Natural Language Processing)

With this Project, we will develop sentiment analysis software using the NLP concept. In this study, we will use the data set obtained from the Kaggle platform, a platform belonging to Google. Thanks to our artificial intelligence software that we will develop in this project, we will be able to automatically extract positive or negative comments from the English IMDB movie reviews that come with this data set. With this project, you will learn the concept of NLP in a very short time without drowning in theory.

– Project #6: Building a Movie Recommendation System

Using the IMDB movie dataset, we will make a software that recommends 5 different movies that are most similar to that movie for any user watching a particular movie. You know, when you watch a movie on NETFLIX, it says the following may also interest you, just like that. While doing this, we will establish a Recommendation System by analyzing the likes of all users in the database who watched and liked the movie.

– Project #7: Predicting Diabetes using Artificial Neural Networks

In this project we are goint to predict whether or not a patient has diabets. We are going to use a well known dataset from Kaggle: Pima Indians Diabetes Database. In this dataset we have some medical test results and statistical information of 768 patients. We will have two different Artificial Neural Network solutions for this project:

• We will build the simplest ANN model using only 1 neuron

• We will build another model using 2 hidden layers and a total of 25 neurons

– Project #8: Image Classification using Convolutional Neural Network and Artificial Neural Network Algorithms (Deep Learning)

We will make a project that automatically recognizes and classifies thousands of different image files using deep learning and artificial neural network algorithms. We will use Tensorflow and Keras libraries to achieve this.

– Project #9: Airline Passenger(Time Series) Prediction using Keras LSTM (Deep Learning)

We will use the Airline Passenger dataset for this project. This dataset provides the monthly total passenger numbers of a US airline from 1949 to 1960. We will produce a solution for this project by using the LSTM model available in Keras, and you will see a good example of how to solve Time Series problems with Deep Learning in general.

– Project #10: San Francisco Crime Geographical Clustering using Machine Learning

In this project, we will perform geographic clustering using Geolocation information (Latitude & Longitude) using a data set created by the SFPD (San Francisco Police Department), which includes crimes committed in the city of San Francisco between 2003-2015. We will also learn to determine the optimal number of clusters (hyperparameter K-value) for this data set using the Elbow method. Then, we will display the geographic coordinates in our clustering results on a Python-based geographic map system. Finally, we will learn how to export this map we created to an HTML file.

– Project #11: Image Classification (ImageNet Library) using Transfer Learning - Keras InceptionResNetV2 (Deep Learning)

Transfer learning uses "knowledge gained in solving a problem" and applies it to a different but related problem. In Transfer Learning, we use a model that has been previously trained on a dataset and includes weights and biases that represent the properties of the dataset it was trained on. In this project, we will use the InceptionResNetV2 model, which has a pre-trained 164-layer advanced architecture and is pre-trained with an ImageNet dataset containing more than 1 million images.

– Project #12: Military Aircraft (Satellite) Imagery Classification using Deep Learning (Custom Datasets)

In this project, we will classify military aircraft images obtained from satellites (F-22 Raptor, Boeing B-52, A-10 Thunderbolt, .. etc.) using Deep Learning algorithms. In this project you will learn to create your own dataset and you will learn to use these customized datasets on pre-trained models.

– Project #13: Sound Signal Processing for Deep Learning using Python (Custom Datasets) (Part - 1/2)

In order to perform Sound Recognition and Classification with Python, the audio files must be in a format that can be used in Deep Learning algorithms. This project is essentially a pre-request project of our next project in our course, “Project#14 - Sound Classification using Deep Learning” Project. In this project we will process sound signals using Mel-Frequency Cepstral Coefficients (MFCC) algorithms and prepare audio for deep learning use. In this project you will learn how to prepare and process your own custom audio dataset for Deep Learning Training and Test operations.

– Project #14: Sound Classification using Deep Learning (Part - 2/2)

We will build a CNN (Convolutional Neural Network) Architecture with three Hidden Layers and 500 neurons in total (125-250-125) using Tensorflow and Keras libraries. We will use the pre-processed sound signals from previous project (Project #13) which has a dataset with a total size of 5.8 GB audio.

Each project will be implemented by Python using Jupyter Notebook. Python source code of each project is included in relevant Udemy course section. You can download source codes for all projects..

This course will cover the following topics:

• Machine Learning (ML)

• Deep Learning (DL)

• Linear Regression

• Customer Segmentation

• Natural Language Processing (NLP)

• Artificial Neural Network (ANN)

• Convolutional Neural Network (CNN)

• Time Series Prediction

• Sentiment Analysis

• Image Classification

• Geographical Clustering

• Sound Signal Processing for Deep Learning Models

• Audio Classification with Deep Learning

Here in this course you will find Artificial intelligence projects for beginners as well as Intermediate/Advanced Level Artificial Intelligence Projects. at the end of the course you will have a clear artificial intelligence definition in your mind and you will get the answer of the question "what is AI ?" or "What is Machine Learning / Deep Learning?"

-----------------------------------------------------------

"When I go visit different companies even at the top Silicon Valley companies, very often I see people trying to apply machine learning algorithms to some problem and sometimes they have been going at for six months. But sometimes when I look at what their doing, I say,I could have told you six months ago that you should be taking a learning algorithm and applying it in like the slightly modified way and your chance of success will have been much higher."

Andrew NG (Professor at Stanford University Department of Computer Science and Department of Electrical Engineering)